YT YouTube2Text

YouTube2Text → Transcripts → Everyone is Cheating (Even the Professors)

Everyone is Cheating (Even the Professors) — Transcript

by Jared Henderson · 4,060 words · 606 segments · language en · Watch on YouTube

Open in transcriber Summarize with AI Download SRT/VTT

Full transcript

0:00This video is sponsored by Private
0:01Internet Access. Stick around to the end
0:03to find out more. A student from Nor
0:04Eastern University demanded her tuition
0:06back after she discovered that her
0:08professor was using Chat GPT to teach
0:10the class. It wasn't fair, she said,
0:12because this professor had banned
0:14students from using AI while he was
0:16using it himself. At another university,
0:17a student got some good news. She had
0:19gotten an A on a paper. So then she went
0:21to go and look at the comments that the
0:22professor had left on it, and she found
0:24that the professor had copy and pasted
0:26an exchange with chat GPT. Not only was
0:28the professor using AI to grade the
0:30paper, but they were sloppy about it,
0:32too. More and more students are using AI
0:34for all of their work. And as one writer
0:35at New York Magazine put it, "It seems
0:37like everyone is cheating their way
0:39through college." The way people tell
0:40this story, AI is ruining education. And
0:42on top of that, Ginz is too lazy to do
0:45any work. But cheating at universities
0:46is nothing new. It's a long-running
0:48issue, and it has to do with the
0:49incentives of education in the United
0:51States. And it even involves the
0:53widespread issue of academic fraud from
0:56professors. We'll explore all of that
0:57today, starting with this latest
0:58problem, AI.
1:02When Open AI was founded, it had a
1:04mission. It wanted to ensure that all of
1:06humanity benefited from artificial
1:09general intelligence. If you read a lot
1:11of the early writing that was coming out
1:12of OpenAI, it sounds rather utopian.
1:15They were saying that they were a
1:16nonprofit. They would be focused on just
1:18doing research and would even be able to
1:19give away that research so that it would
1:21benefit the world. All of humanity
1:23should benefit from this thing they
1:24wanted to create. true AI. Open AAI
1:27might be a nonprofit, but within OpenAI,
1:30there's a for-profit company, and it's
1:31this for-profit company that tries to
1:33bring products to market. And OpenAI is
1:35now worth something like $300 billion.
1:37And that's because people really want
1:39what they're selling. Every major
1:40artificial intelligence company now
1:42heavily markets themselves towards
1:44students, and it's definitely working.
1:45Nearly half of students and teachers at
1:47the K through 12 level say that they use
1:49AI, and nearly every college student
1:52uses AI in some capacity. The result is
1:54that we have issues of plagiarism,
1:56cheating, and fraud. AI can write papers
1:59quickly. It can make it easy to pass a
2:00class, and it frees up a student's time
2:03so that they can do stuff that they'd
2:04rather be doing. Whether that's working
2:06a job or thinking about their career
2:08prospects or just, you know, going out
2:10and drinking or something. There are now
2:11apps, websites, other services, all
2:13dedicated to helping students use AI to
2:16cheat and pass their classes. Professors
2:18who want to avoid cheating have to find
2:19workarounds. They'll sneak in extra
2:21words into assignments that are designed
2:23to mislead AI. Or they'll demand that
2:25students take exams with blue books,
2:27like literally writing a test with, you
2:29know, by hand with a pencil or a pen.
2:31This might solve part of the problem,
2:32but when it comes to things like
2:34take-home assignments or extended essays
2:36or even short reading responses, well,
2:38the overwhelming odds are that students
2:40are using AI to do basically all of that
2:43work. The consequences of this are
2:45pretty dire. One professor at Cal State,
2:47Chico, said that massive numbers of
2:48students are going to graduate from
2:50university and they're going to enter
2:51the work and they will be essentially
2:52illiterate. I think the concern is that
2:54whatever skills we think you're supposed
2:56to develop when you go to college, when
2:57you go to university, aren't being
2:59developed at all. That's because more
3:01students are finding out they don't need
3:03to read or write or have any original
3:05thoughts in order to get their degrees.
3:07Some universities like Ohio State have
3:09decided that the solution is to just
3:11embrace AI completely. Ohio State
3:14actually says that every student will be
3:15encouraged to use AI in every class from
3:18now on, but most universities simply
3:20don't know what to do. The philosopher
3:22Megan Fritz, who teaches at the
3:23University of Arkansas at Little Rock,
3:25wrote an article about this. Fritz sat
3:28on several university committees that
3:29were trying to come up with AI policies
3:32that could help, you know, solve some of
3:33these problems at her school. And they
3:36got nowhere. They tried to develop
3:37policies for departments, but they
3:39couldn't keep up with the progress from
3:41various AI companies. the technology was
3:43just moving too quickly. And in an essay
3:45for the point, she wrote that the
3:47biggest difficulty was determining what
3:49the goal of an education was supposed to
3:51be. And if you can't answer that
3:53question, then you probably can't answer
3:55any of the questions about AI and
3:58cheating.
4:01So in his book, The Case against
4:02Education, the economist Brian Kaplan
4:05makes the case that Americans spend too
4:08much time and too much money on
4:10education in general. I would argue
4:12against the main thrust of Kaplan's
4:14book, I think, but he does make an early
4:16point that's really important for this
4:18discussion. Much of modern education is
4:20really not about passing on skills or
4:22learning things. It's simply about
4:24getting the degree at the end of the
4:26day. And that's because a degree signals
4:28certain facts about you. And Kaplan says
4:30that those include intelligence,
4:31conscientiousness, and conformity. Many
4:34employers will require a college degree
4:35for a high-paying job, but they're not
4:37necessarily looking for specific skills
4:39that you acquired in the course of
4:40getting that degree. An art history
4:42major might end up working on Wall
4:43Street. After I got a PhD in philosophy,
4:45I went into tech. And it's not like my
4:47PhD in philosophy was directly setting
4:49me up to go work in tech. Many people
4:50get degrees in things like
4:52communications or psychology. And they
4:53get jobs that do not require you to know
4:55anything about theories of communication
4:57or facts about the human mind. By the
4:58way, my new puppy Apollo is joining us
5:01for this video because he really likes
5:02to be held. If the skills you acquire in
5:05the course of getting a college degree
5:06don't actually matter, then why do
5:08employers require one? It's because a
5:09college degree does usually signal some
5:12facts about the job seeker. It just has
5:15very little to do with the content of
5:16your education. It basically says that
5:18you were able to survive for 4 years at
5:20an institution that has kind of arcane
5:22and arbitrary rules and that you were
5:25good enough at following those rules and
5:26smart enough to navigate them that you
5:28could get a degree at the end. And guess
5:30what? Working at a large corporation is
5:33going to be an experience of surviving
5:35an arcane and obscure institution with
5:38weird rules. Kaplan's theory doesn't
5:39really account for specialist degrees
5:41like engineering, but we can set those
5:42aside because those are a minority of
5:44degrees. I think deep down actually many
5:46students realize that this is true. They
5:48look at many of the classes they have to
5:50take in college and they think why does
5:52any of this matter? All I'm looking to
5:53do is to just get a better job at the
5:55end of this. If you're going to college
5:56to increase your career prospects or to
5:59climb the social ladder, well, you know,
6:01philosophy 101 isn't necessarily going
6:03to help you do that. And if you're a
6:04student who understands this, then using
6:05AI to cheat isn't just reasonable. It
6:08might be the rational thing to do. At
6:10least it is within your self-interest.
6:11You're not there to wax philosophical
6:14about continism versus utilitarianism or
6:16to go and write an essay about Charles
6:18Dickens. I wish more people wanted to do
6:20those things, but it's not why most
6:22people go to school. Kaplan really is
6:23just thinking about the incentives of
6:25getting a college degree. And I think if
6:27there's like a big thesis for this
6:28video, it's that incentives matter. To
6:30make Cap's point a little bit clearer,
6:31you could look at the phenomenon of
6:33diploma mills. These are institutions
6:36that set themselves up like colleges and
6:38universities. They are often for-profit
6:39and unacredited. And in exchange for
6:42some money and very little work, you can
6:44get a college degree. It's even a way to
6:46get a PhD in some cases. Some schools
6:48will simply give you the degree if you
6:50pay them enough money. That's the most
6:51egregious example. But other schools
6:52will simply just make it very easy for
6:55you to get the degree. They'll count
6:56work experience as credit, for instance.
6:59Um, and you'll be able to get out of
7:01most of your coursework. So, you just
7:03take a couple of classes, you write
7:04120page paper, and bam, you have a
7:06master's degree from online university.
7:09And if all you need on your resume is
7:11just a degree from some school so that
7:14some AI filter at some company will let
7:17you through so that a human being will
7:19actually just read your resume. Well,
7:22then it kind of makes sense to go and
7:24get a degree. Even though these things
7:25are, you know, scam, not having a degree
7:28is a problem if you're looking for a
7:29job. A fake degree solves that problem.
7:32and you don't have to learn anything.
7:34Perhaps the tragedy of what is happening
7:35at many universities where students
7:37could just use AI to cheat on everything
7:38is that kind of all education is
7:41eventually just going to be fake and
7:43made up because nobody actually has to
7:44learn anything to get a degree.
7:48When I read Kaplan's book, I immediately
7:50saw a connection to another issue in
7:52academia, the widespread presence of
7:55academic fraud. This is research fraud
7:57that's committed by professors.
7:59Recently, there was the case of a
8:00Harvard Business School professor,
8:02Franchesco Gino. Her tenure was revoked,
8:04which almost never happens, and she was
8:06fired from Harvard after an
8:08investigation found significant evidence
8:10of fraud in her research. It seems she
8:12had used fake data in multiple studies.
8:14And some of the papers where she faked
8:16this data was research on authenticity
8:18and honesty. It's very rare for tenure
8:21academics to ever be fired. This, if you
8:23can get tenure, it is basically a
8:25guarantee of a job for life. And that
8:27should show you just how serious these
8:28infractions were. The investigation into
8:30Gino's work started several years ago,
8:32and it actually began with some science
8:33bloggers. That's where a lot of academic
8:35fraud seems to be discovered. These
8:37bloggers found significant issues in at
8:39least four of Gino's published papers.
8:41Gino tried to say that she was a victim
8:42of defamation and gender discrimination
8:45and that her privacy had been invaded.
8:46Then she sued Harvard. She sued the dean
8:48of the Harvard Business School, her
8:50boss, and she sued those bloggers. But
8:52that case was thrown out by a judge. And
8:54now Harvard Business School has
8:55determined that she really did fake data
8:57and now she's lost her job. But Gino is
9:00not alone in committing academic fraud.
9:02Throughout 2024, there were many
9:04high-profile academics who were found
9:06guilty of plagiarism or as they
9:08sometimes put it, inadequate citation,
9:11including a former professor of Harvard.
9:14And just recently, a PhD student at MIT
9:16in the economics program had a paper
9:19pulled and I believe is no longer a
9:21student at MIT after the university
9:23investigated it. And they went on to say
9:24they had no confidence in the veracity
9:26of the research. Because of privacy laws
9:29and university policies, MIT really
9:31can't say much beyond that. They don't
9:32even use the student's name in that
9:34announcement. Though they give the title
9:36of the paper, and if you were to look
9:37for the title, you would easily find the
9:38author's name. This might seem fairly
9:40minor because students cheat on papers
9:42all the time. However, this paper was
9:43scheduled to be published in a good
9:45economics journal, and it had already
9:46been posted as a preprint on the website
9:49archive. That's a website where
9:50scientists post copies of their research
9:53so that it's more freely accessible. In
9:55this case, the problem was that the
9:57research was fraudulent. If the paper
9:58had been legitimate, it would have made
10:00this student's career because it had a
10:02really provocative conclusion. It seemed
10:04to show that using AI actually boosted a
10:07firm's research productivity. So
10:09incorporating AI would actually make
10:11employees more productive. This is a
10:13result that many people who are big fans
10:14of AI would love to see. It would seem
10:17to be a great response that like
10:18skeptics of generative AI. People were
10:20naturally interested in this paper and
10:22it was getting a lot of buzz from the
10:23moment it was posted on archive. And if
10:25you're a graduate student, that's the
10:27sort of paper that you would want to
10:29publish as you go into the job market.
10:30It's the sort of paper that would
10:32identify you as a young and brilliant
10:33scholar who was going to produce great
10:35research and you would definitely get a
10:37job. You could be set for life. based on
10:38a paper like that. And sometimes this
10:40academic fraud has real world
10:42consequences. One of the most famous
10:44academic frauds is Brian Wansync.
10:46Wansync was a nutrition scientist at
10:48Cornell and he's actually responsible
10:51for some studies that entered the public
10:53consciousness. You might believe some of
10:55the things that was initially discovered
10:57that's discovered in quotes by Wansync
11:00and his team. Have you ever heard that
11:01shopping on an empty stomach is a bad
11:03idea because you'll buy more food or
11:05maybe you'll buy more junk food? that
11:07was originally from a wansing study and
11:09I believe that study has been
11:10discredited along with much of his other
11:12research. He was also responsible for
11:13that idea that the size of your plate
11:15determined how much you wanted to eat at
11:17a meal. In 2017 it was discovered that
11:19Wansync had in some cases fabricated
11:22data and in other cases he had done
11:24misleading analyses of the data to try
11:26and support his claim. The the simplest
11:28way to describe that though is that he
11:30committed fraud. He was actually found
11:31out because he wrote a blog post about
11:33his own methodology. And he admitted
11:35that sometimes he would have graduate
11:36students go over experiments that seem
11:38to be initial failures and go looking
11:41for interesting results in the data. And
11:43this is a classic sort of research
11:45misconduct. And Lancync like Gino was
11:47only discovered because some science
11:49bloggers got interested in the issue.
11:51They started looking into it and then
11:53once they found something interesting
11:54they kept digging. When Cornell reviewed
11:56his research, they found many problems,
11:58including misreporting research data,
12:01problematic statistical techniques,
12:03failure to properly document and
12:04preserve research results, and
12:06inappropriate authorship. When we're
12:08talking about student misconduct at
12:10universities, like using AI to cheat in
12:12all their classes, I think it's also
12:14important to look at professors and ask
12:16questions about why do so many
12:18professors commit research fraud? Or you
12:20could ask the more general question,
12:21just why do they lie?
12:26We like to think of scientists as noble
12:28individuals in pursuit of truth. How are
12:30they tempted to abandon this pursuit of
12:32truth and then go and commit fraud? One
12:35of the best writers on this subject is
12:36Liam Kofi Bright. He's a philosopher at
12:38the London School of Economics and he
12:40studies the social epistemology of
12:42science. In other words, he studies how
12:44the institutions and group dynamics of
12:46science affect scientific progress. And
12:49in a paper, why do scientists lie? Brite
12:51points out that if you have a PhD in
12:53physics or some other science, you could
12:55actually go and get rich by leaving
12:57academia and entering industry. If
12:59you're looking to get rich as a
13:00scientist, you're better off just
13:02leaving academia. And that would suggest
13:04that most professors don't commit fraud
13:06because they're greedy. If they were
13:08really greedy and just wanted more
13:09money, they would just quit their
13:10university jobs. Wright instead defends
13:12an idea that's originally due to Robert
13:14Mertin. Mertin theorized in the 1960s
13:17that scientists are what we could call
13:19credit seekers. Another term that Bright
13:21and Martin use is honor. They seek
13:23status in the scientific community. The
13:26way we have arranged our institutions
13:27and you achieve status by producing
13:29research and we typically want novelty.
13:31So we want the results to be new and
13:34volume of output. Most academic
13:35researchers go through something called
13:36a tenure review at some point in their
13:39career. This is where academics evaluate
13:41your work and they decide whether or not
13:43you should be granted tenure. And if you
13:44get tenure at a school, you are
13:46basically guaranteed a job for life.
13:47You're not going to become fabulously
13:48wealthy this way, but it essentially
13:50guarantees an upper middle class
13:52existence. And if you fail your tenure
13:53review, typically you're fired. And to
13:56pass your tenure review, you prove that
13:57you're a good researcher and and two of
13:59the major metrics we use to judge this
14:01are the number of articles or books that
14:03you've published and how often you're
14:05cited by other people. So essentially,
14:07we're asking, do you produce a lot of
14:08research? And is that research
14:10interesting enough that people keep
14:11talking about it when they do research?
14:13Turns out that both of these metrics can
14:14be gamed. There are journals that you
14:16can pay to publish your academic papers.
14:18It's fairly easy to get caught doing
14:20this, but some academics still do it.
14:22Another way, though, is to commit fraud.
14:24And most academics who commit research
14:26fraud probably don't think they're doing
14:27anything that bad. And they're not going
14:29to do things that are as egregious as
14:31some of the people that we've talked
14:32about today, like where they just fake
14:33data and make things up. So instead,
14:35they're fudging statistics. They
14:37slightly misdescribe the evidence. They
14:39might even believe that they're doing
14:41nothing wrong. When you're properly
14:42incentivized, you can deceive yourself.
14:44If you want to then get cited more,
14:46well, you can pay for that, too. You can
14:47still get caught. So, again, if you want
14:49to avoid scandal and boost your
14:51citations, you just need to publish new
14:53and interesting ideas. Take a look at
14:54how Brite puts it. Scientists want and
14:57need credit for new results to ensure
14:59their results are new, i.e. novel enough
15:01to have a plausible claim to
15:02establishing priority. Scientists have
15:04to do their work sufficiently quickly
15:06that nobody else beats them to the
15:08punch. So because these scientists have
15:10to work fast, they are incentivized to
15:12either be sloppy or to commit fraud.
15:15See, that's the incentive. Researchers
15:17commit fraud because they are rewarded
15:19for publishing often and for publishing
15:21big results and for appearing to be
15:22geniuses. So they they fudge the
15:24numbers. They sometimes outright lie and
15:26they fabricate data. Other times you
15:28plagiarize where you just steal research
15:30from other people. In other words, you
15:32find a way to cheat. I think when
15:34students decide to cheat in college,
15:36essentially what they're trying to do is
15:38game the system. They want the economic
15:40or lifestyle benefits of an education,
15:42but they don't want to have to put in
15:43the work. Human beings are often lazy.
15:45They look for shortcuts. It's a normal
15:47way to behave. It might not be an
15:49admirable way to behave, but it's at
15:51least normal. And the academics who
15:53commit fraud are doing roughly the same
15:55thing. They want the benefits of having
15:57done this great research, but they look
15:59for shortcuts. You look at both of those
16:01cases, there's one common lesson simply
16:03that incentives matter. How you
16:05structure an institution matters and how
16:08an institution is structured and what it
16:10rewards matters. There aren't going to
16:12be quick fixes to either problem. You
16:15know, part of the problem is you have to
16:16convince people that being truthful or
16:18doing your homework or really thoroughly
16:20going over all your data matters. And
16:23essentially what you're doing is
16:24convincing people to spend more of their
16:25time doing things that probably they
16:27don't really want to do. And and this is
16:29I think the big challenge that
16:32universities around the country are
16:34going to be facing. I've talked many
16:35times in this channel about the ways
16:37that universities have sort of lost
16:39their way. They are beholden to a
16:40certain economic model of education
16:42which can make it difficult to see why
16:44actually getting an education in the
16:45first place matters and we're seeing
16:47more attacks on higher education from
16:49various parts of the government. A day
16:52or two before recording this, I saw the
16:54news that Indiana University, which is a
16:56premier state university that's known
16:58for its language programs, just cut many
17:01of those programs due to low enrollment.
17:03There's a new state law that says that
17:04if you don't have a certain number of
17:05graduates per year, the major and the
17:07PhD program and everything just go away.
17:10And and so now they've just gutted some
17:12of the best language departments in the
17:13country. When it comes to education in
17:15the United States, I think we're at an
17:17inflection point. Universities had to
17:18figure out what they want to offer to
17:19the world, what they want to offer to
17:21students and what they want to be going
17:24into the future.
17:26People might know that I used to be an
17:27academic and one of the most difficult
17:29things about leaving academia and
17:31becoming an independent researcher was
17:33that I lost access to institutional
17:36libraries and that means I don't have
17:37access to things like scholarly
17:38journals. So sometimes I end up going
17:40down rabbit holes on the internet trying
17:42to find the right PDFs. And whenever I
17:44do that, I want to make sure that I'm
17:45using a VPN. And in particular, I use
17:47Private Internet Access, and they're
17:49sponsoring this video. A VPN essentially
17:51is a way for you to browse the internet
17:53more securely. It hides your IP address
17:55and safeguards your internet connection
17:56through an encrypted tunnel. This way,
17:58it shields your digital life and keeps
17:59your personal information from people
18:01who might want to exploit it. Think
18:02people like hackers or internet service
18:04providers or even governments. On top of
18:06that, many services and websites have
18:08content that can only be accessed if you
18:10have the right IP address. Basically,
18:12you have to be accessing it from a
18:14certain country. There are shows or
18:15movies you can't watch on Amazon or
18:17Netflix or the BBC unless you're
18:19accessing it from the right place. I
18:21have a hard time watching, you know, the
18:22latest Louis Thorough documentary from
18:24the BBC because I'm not in the UK. A VPN
18:27lets you choose your digital location.
18:29In fact, with PIA, you can choose from
18:3191 countries. And in the US, they have
18:34servers in all 50 states. PIA works with
18:36all major streaming services and they
18:38are one of the few VPNs that supports
18:40P2P file sharing and you can use it on
18:43Windows, Mac OS, Android, Linux, and
18:46iOS. With a single subscription, I am
18:48able to protect all of my devices at the
18:50same time. PIA is the world's most
18:53transparent VPN provider. They never
18:55record or store user data, and their no
18:57logs policy has been proven multiple
18:59times in court. And for an extra level
19:01of confirmation, this was confirmed by a
19:03third-party audit from Deote. If you use
19:05my link, viewers can get an 83% discount
19:08on PIA. That's just $2.3 a month, and
19:11you get four bonus months for free. You
19:13can use the link in the description, or
19:14you can use the QR code that's on the
19:16screen right now. Signing up is
19:17risk-free, allowing a 30-day money back
19:19guarantee, and they have 24/7 support
19:22available to answer any of your
19:23questions.

About this transcript

This page contains the full transcript of Everyone is Cheating (Even the Professors) by Jared Henderson, generated from the public captions YouTube serves with the video. The transcript has 4,060 words across 606 segments, with the original timestamps preserved so you can click any line to jump to that moment in the embedded player.

What you can do with it

Use the transcript to take notes, quote the speaker, build a study guide, generate a summary with ChatGPT or Claude via the YouTube Summary tool, or export it as a timed subtitle file with YouTube to SRT. You can also re-open it in the transcriber to translate the transcript into 100+ languages.

Free YouTube transcript tool

YouTube2Text is a free YouTube transcript generator — no signup, no daily limit. Paste any YouTube link and get the full transcript instantly, with timestamps, click-to-jump, translation to 100+ languages, AI prompts for ChatGPT, Claude, and Gemini, and exports to TXT, SRT, VTT, or Markdown.